Why is there such a gap between the evidence base for effective drug treatment and its dissemination in the community? Among behavioral interventions, Motivational Interviewing (MI) should be well positioned for high-impact dissemination: There is an international organization devoted to the training and dissemination of MI (The Motivational Interviewing Network of Trainers [MINT]), and MI developers have been at the forefront of research on dissemination. In addition, MI mechanisms are well specified and many are considered basic building blocks of behavioral interventions (e.g., empathy, use of reflections and open questions).1 Empathy, a central MI component, has a strong research tradition in basic science and is among the most consistent predictors of positive outcome in psychotherapy across treatments and disorders.2 At present, quality control in MI relies on ongoing supervision and feedback to therapists in the community. However, measuring therapist behavior is synonymous with behavioral coding by humans, which is not feasible in community settings3. Accordingly, there is no feasible mechanism for evaluating the practice of evidenced based treatments in clinical settings. Advances in linguistic processing have brought a technology for behavioral coding within reach. Specifically, in the basic science and engineering fields, there are tools that combine acoustic and semantic information into predictive models of human-defined knowledge (e.g., empathy codes).4-8 The current proposal brings together a highly interdisciplinary team to produce a viable technology for generating feedback on therapist empathy without human intervention. This technology may eventually be packaged and disseminated in the community. As an initial step, we will apply a sequence of linguistic processing approaches to develop a process for predicting therapist empathy from audio files obtained from MI sessions (AIM 1).9 Next, we will run a feasibility study to determine the performance of the coding technology with new data obtained from MI sessions conducted by novice and experienced therapists (AIM 2). During this evaluation we will troubleshoot issues related to audio quality and speech recognition as well as the process of feedback. In addition, we will evaluate the agreement of computer-based codes with human coding of new sessions. Drug abuse and dependence represents an incredible societal burden. The current research will develop an innovative process for enhancing the ability disseminate and maintain quality control in the behavioral treatment of addictions.